The long-term objectives of this research are: (1) to identify dynamic and geometric features of the coronary vasculature that favor atherosclerotic development and can be used prognostically and to identify individuals at risk, and (2) to understand the role of mechanical factors in the pathobiology and natural history of atherosclerosis. To achieve the first objective, (a) the normal variability in human coronary artery dynamics and geometry will be defined, and (b) relationships will be sought between the dynamic geometry of diseased human coronary arteries and the distribution and progression of pathology in these vessels. The second objective will be furthered by (a) developing models of intramural stress distribution for testing against measured in vivo wall strain, and (b) correlating the dynamic geometry of the coronary arteries of apo-E deficient mice against tissue histology and gene expression. Biplane coronary cineangiograms of 140 patients with angiographically normal vessels will be processed to reconstruct the time-dependent three-dimensional (3-D) course of the medial axes of selected segments of the left anterior descending and right coronary artery trees. Parameters describing the dynamic geometry of the segments will be obtained from the axes using objective computer algorithms, and their normal distributions in the population will be defined. For an additional cohort of 60 patients with evident disease, angiograms will be complemented by intravascular ultrasound (IVUS) records of the same segments, which will be processed to yield detailed measurements of vessel morphometry and 3-D strain collocated to the vessel axes. In some cases, phasic pressure will also be acquired. Relations will be sought between the parameters derived from the angiograms and from the IVUS records, taking into account variability in the traditional risk factors for atherosclerosis. These data will also be used to develop and test models of stress distribution and adaptation in the vessel wall. The hearts of 36 apo-E deficient and 8 wild type control mice at four ages will be exposed and viewed in biplane. The biplane images will be processed for dynamic geometry parameters similarly to the clinical images, and relationships will be sought between these parameters and the morphometry, histology and gene expression profiles of the tissue.